Time for a sanity check
So I've nearly got my 1st stator panel completed (been busy with unbelievable stuff the past few months).
I have the wires on both ends soldered together and it is time to snip them into the physical segments.
So here is what I am planning.....
There are 12 electrical segments, 23 physical segments. That means one physical segment in the middle with 11 physical segments on both side.
Given the thread spacing (13 wires per inch) that equals 8 wires per physical segment and about .6 inches wide (1.2ish electrical).
On the bolserst's spreadsheet, config 2, it shows an input of 5,5,.... wires on the input to the esl seg ui app..... guessing this doesn't me what I think it means since I would expect 8,8,......
I am thinking of just stretching the diaphram to 2% and roll with it. I figure it will relax over time with the Fs drifting down.... I am expecting my audyssey room/eq correction to tame down any resonance at Fs if there is any... The active crossover I am using is 24dB 4th order linkowitz/reily (@250hz).
On the transformer front. I plan to use (and have ordered) 230/6v transformers.... 4 per panel.
In this config, that gives me a step of 153X. So here is my current tinkerings with the bias and step up config. One thing I am unsure of is, if I upgrade my amp to 200W (etc...) the input voltage will go "red". If I crank up the amp beyond 28V output, will my ESL panels explode into flames?
e0: 8.85E-12 N-m2/C2
c: 340.00 m/s
Emax: 4,000 V/mm
1 ==> h: 46.50 (in) height
2 ==> w: 14.10 (in) width
3 ==> d: 0.0625 (in) gap
4 ==> r: 15.0000 (ft) distance
A: 0.4230 (m^2) area (panel)
h: 1.1811 (m) height
d: 0.0016 (m) gap
Vpol: 3,175 (VDC) Vbias (optimal)
Vsig: 6,350 (Vpeak) Vsignal (max/optimal)
5 ==> N: 12 # sections
6 ==> fL: 250.00 (Hz) LFbreak pt
fH: 144000.00 (Hz) HFbreak pt
R: 44.99 (Kohms) Feed Resistance
Ctot: 1179.07 (pF) Cpanel
C: 98.26 (pF) Csection
w(sec): 1.18 (in) width section
28.19 (Vrms) input limit
7 ==> Vin: 28.00 (Vrms) amplifier voltage
8 ==> Step-Up: 153.0 step-up ratio
9 ==> Vbias 3,300.0 (VDC) bias voltage
10 ==> Line Source: Finite Type of Line Source
Vstators 4,284 (Vrms) stator-to-stator voltage
6,058 (Vpeak) stator-to-stator voltage
11 ==> Fs: 100.00 (Hz) Resonance Frequency
12 ==> Q: 10.0 Resonance Q
I considered only using 3 transformers or using the 1 of them in 12V series to lower down the step up. Not sure if either is a good idea.
All of this sound good? Anything I am missing or should reconsider?
So I've nearly got my 1st stator panel completed (been busy with unbelievable stuff the past few months).
I have the wires on both ends soldered together and it is time to snip them into the physical segments.
So here is what I am planning.....
There are 12 electrical segments, 23 physical segments. That means one physical segment in the middle with 11 physical segments on both side.
Given the thread spacing (13 wires per inch) that equals 8 wires per physical segment and about .6 inches wide (1.2ish electrical).
On the bolserst's spreadsheet, config 2, it shows an input of 5,5,.... wires on the input to the esl seg ui app..... guessing this doesn't me what I think it means since I would expect 8,8,......
I am thinking of just stretching the diaphram to 2% and roll with it. I figure it will relax over time with the Fs drifting down.... I am expecting my audyssey room/eq correction to tame down any resonance at Fs if there is any... The active crossover I am using is 24dB 4th order linkowitz/reily (@250hz).
On the transformer front. I plan to use (and have ordered) 230/6v transformers.... 4 per panel.
In this config, that gives me a step of 153X. So here is my current tinkerings with the bias and step up config. One thing I am unsure of is, if I upgrade my amp to 200W (etc...) the input voltage will go "red". If I crank up the amp beyond 28V output, will my ESL panels explode into flames?
e0: 8.85E-12 N-m2/C2
c: 340.00 m/s
Emax: 4,000 V/mm
1 ==> h: 46.50 (in) height
2 ==> w: 14.10 (in) width
3 ==> d: 0.0625 (in) gap
4 ==> r: 15.0000 (ft) distance
A: 0.4230 (m^2) area (panel)
h: 1.1811 (m) height
d: 0.0016 (m) gap
Vpol: 3,175 (VDC) Vbias (optimal)
Vsig: 6,350 (Vpeak) Vsignal (max/optimal)
5 ==> N: 12 # sections
6 ==> fL: 250.00 (Hz) LFbreak pt
fH: 144000.00 (Hz) HFbreak pt
R: 44.99 (Kohms) Feed Resistance
Ctot: 1179.07 (pF) Cpanel
C: 98.26 (pF) Csection
w(sec): 1.18 (in) width section
28.19 (Vrms) input limit
7 ==> Vin: 28.00 (Vrms) amplifier voltage
8 ==> Step-Up: 153.0 step-up ratio
9 ==> Vbias 3,300.0 (VDC) bias voltage
10 ==> Line Source: Finite Type of Line Source
Vstators 4,284 (Vrms) stator-to-stator voltage
6,058 (Vpeak) stator-to-stator voltage
11 ==> Fs: 100.00 (Hz) Resonance Frequency
12 ==> Q: 10.0 Resonance Q
I considered only using 3 transformers or using the 1 of them in 12V series to lower down the step up. Not sure if either is a good idea.
All of this sound good? Anything I am missing or should reconsider?
Last edited:
Hi Bengel
The ESLs with not explode into flames 😀
When you first fire it up, you might find that there is a little bit of snap crackle pop as tiny bits of dust get blown away, but after that, and whenever you power it up afterwards, it should be whisper quiet.
If your HV supply is adjustable, start with a low voltage, say 1.5 kV and work your way up to design voltage. It is easier to diagnose problems if there is a little sizzle happening rather than a large bang and your HV is gone and you don't know where to look for the problem. Power up one speaker at a time for same reasons.
Similarly for the audio side. Put the set the step up ratio to 75X to start with by putting half of the 6V winding in series rather than all in parallel. Once you are happy that all is working well, increase the step-up ratio. to 150X
If you have the crossover working, you should be able to get the ESLs very loud before the transformers saturate. So long as your amp has output current protection, there should be no problem. In any case make sure there is a series resistance of at least 1 ohm in the primary circuit to load the amplifier - start with a higher resistance perhaps. - low wattage resistors could serve as a fuse. If you use several ohms, you should expect the upper end of the treble to disappear.
Remember to keep one hand in your pocket when playing (seriously - it will avoid electrical currents traversing your heart if you do get a shock) , and disconnect and discharge the HV before you put hands anywhere near the ESLs - be patient and very careful. The HV is high impedance (you should have 100 M or similar in series connection to the panel) so it will bite hard and really hurt, but only be up to 1 or 2 on the crocodile scale. The audio is a different question, it is low impedance and potentially lethal. NEVER have your hands in the ESLs when audio is live.
Good luck
The ESLs with not explode into flames 😀
When you first fire it up, you might find that there is a little bit of snap crackle pop as tiny bits of dust get blown away, but after that, and whenever you power it up afterwards, it should be whisper quiet.
If your HV supply is adjustable, start with a low voltage, say 1.5 kV and work your way up to design voltage. It is easier to diagnose problems if there is a little sizzle happening rather than a large bang and your HV is gone and you don't know where to look for the problem. Power up one speaker at a time for same reasons.
Similarly for the audio side. Put the set the step up ratio to 75X to start with by putting half of the 6V winding in series rather than all in parallel. Once you are happy that all is working well, increase the step-up ratio. to 150X
If you have the crossover working, you should be able to get the ESLs very loud before the transformers saturate. So long as your amp has output current protection, there should be no problem. In any case make sure there is a series resistance of at least 1 ohm in the primary circuit to load the amplifier - start with a higher resistance perhaps. - low wattage resistors could serve as a fuse. If you use several ohms, you should expect the upper end of the treble to disappear.
Remember to keep one hand in your pocket when playing (seriously - it will avoid electrical currents traversing your heart if you do get a shock) , and disconnect and discharge the HV before you put hands anywhere near the ESLs - be patient and very careful. The HV is high impedance (you should have 100 M or similar in series connection to the panel) so it will bite hard and really hurt, but only be up to 1 or 2 on the crocodile scale. The audio is a different question, it is low impedance and potentially lethal. NEVER have your hands in the ESLs when audio is live.
Good luck
I was going to ask about this but since you brought it up 🙂. What are the pros/cons with putting a resistor in series there?
My amp does have short circuit protection (I've "used it" before 🙂 ) are there any other reasons to have it? I thought I read something about it helping smooth out the impedance curve seen by the amplifier.
Hi bengel
Bunch of things
DC Offsets: most amps have a dc offset on the output, ideally only a few mV. A small resistance ensures that there is a small dc current flowing and that the core is not partially magnetised (which would cause early saturation for some signals).
Distortion: The magnetising current for the transformer core is extremely non-linear. Typically, the magnetising impedance looks like the primary inductance of the transformer in parallel with a non-linear resistance that accounts for core losses. A good rule of thumb for silicon-steel cores is that 40% of the magnetising current is non linear at full signal. Any resistance in series with the primary will cause a non-linear voltage to appear in series with the amplifier output. So, ideally the transformer should have a high primary inductance and there should be zero series resistance. The distortion current is approx. 0.4Vin /(2.Pi.f.L). With 1 ohm in series with 1H inductance, the distortion will be 1/(0.4.2.Pi.f.L) or about 0.1% at 50Hz, and falling in direct proportion to frequency. Human ears are not sensitive to low frequency distortion so this should not be a problem.
Saturation: The steel cores only have a high permeability up to a certain magnetic field strength. Above that, the permeability falls, inductance drops hugely, and the magnetising current races. Your power transformers are designed to be just below this point with 50 Hz and 6Vac applied to the secondary windings, or 230Vac applied to the primary windings - same requirement. The saturation voltage scales with frequency, so at 250 Hz, they will cope with 30Vac before saturating. At saturation, the poor performance of the core will clip the audio waveform, and the amplifier current will race. A series resistance will introduce extra distortion, but also limit the amplifier current a bit - less chance of smoke being lost.
Frequency response: As I mentioned in a previous post , the combination of ESL, transformer, and amplifier inductance gives the system a second -order low pass response, with a cutoff frequency approximately equal to the resonant frequency of the loaded transformer. Altering the resistance of the circuit adjusts the Q of the resonance. If you have a high resistance in the primary, it will lower the Q and the response will begin to fall at the highest frequencies - hence loss of treble. The segmented ESL has sufficient resistance to damp the Q to about 1.5, a small additional series resistance of less than an ohm is normally about right to pull the Q down below 1 giving the flattest response -if you can measure the frequency response of the system when it is set up, you can tweak the series resistance to flatten the top end as required.
more compromises ...
regards
Rod
Bunch of things
DC Offsets: most amps have a dc offset on the output, ideally only a few mV. A small resistance ensures that there is a small dc current flowing and that the core is not partially magnetised (which would cause early saturation for some signals).
Distortion: The magnetising current for the transformer core is extremely non-linear. Typically, the magnetising impedance looks like the primary inductance of the transformer in parallel with a non-linear resistance that accounts for core losses. A good rule of thumb for silicon-steel cores is that 40% of the magnetising current is non linear at full signal. Any resistance in series with the primary will cause a non-linear voltage to appear in series with the amplifier output. So, ideally the transformer should have a high primary inductance and there should be zero series resistance. The distortion current is approx. 0.4Vin /(2.Pi.f.L). With 1 ohm in series with 1H inductance, the distortion will be 1/(0.4.2.Pi.f.L) or about 0.1% at 50Hz, and falling in direct proportion to frequency. Human ears are not sensitive to low frequency distortion so this should not be a problem.
Saturation: The steel cores only have a high permeability up to a certain magnetic field strength. Above that, the permeability falls, inductance drops hugely, and the magnetising current races. Your power transformers are designed to be just below this point with 50 Hz and 6Vac applied to the secondary windings, or 230Vac applied to the primary windings - same requirement. The saturation voltage scales with frequency, so at 250 Hz, they will cope with 30Vac before saturating. At saturation, the poor performance of the core will clip the audio waveform, and the amplifier current will race. A series resistance will introduce extra distortion, but also limit the amplifier current a bit - less chance of smoke being lost.
Frequency response: As I mentioned in a previous post , the combination of ESL, transformer, and amplifier inductance gives the system a second -order low pass response, with a cutoff frequency approximately equal to the resonant frequency of the loaded transformer. Altering the resistance of the circuit adjusts the Q of the resonance. If you have a high resistance in the primary, it will lower the Q and the response will begin to fall at the highest frequencies - hence loss of treble. The segmented ESL has sufficient resistance to damp the Q to about 1.5, a small additional series resistance of less than an ohm is normally about right to pull the Q down below 1 giving the flattest response -if you can measure the frequency response of the system when it is set up, you can tweak the series resistance to flatten the top end as required.
more compromises ...
regards
Rod
BTW, do you mean 1uH inductance?
When you say falling here, do you mean the distortion will fall as frequency increases? i.e. distortion is .1% at 50Hz but it gets "better" from there up to 20khz...
So given they will saturate at 30Vac, I am limited to a max output of 28V from my amp to avoid clipping at the lower end of my panel frequency range?
How does the saturation voltage relate to the size of the core? I would presume bigger cores (i.e. 50VA transformers) would saturate at high voltages.....
But in general, yes, put a 1 ohm or maybe slightly less resistor in series with it 🙂.
Hi Bengel
No, I mean 1 H inductance. This is the impedance that appears in parallel with the reflected load impedance. This tells you how much current the amplifier must provide to generate magnetic field in the core, in addition to the reflected load current.
Yes, distortion falls with frequency 0.01% at 500 Hz, etc...
No you are not limited to 30 Vac. The audio signal comprises many different signals at different frequencies, its the lowest frequencies that determine whether it saturates or not- Which is why it is important that the crossover removes lower frequencies. Remember saturation voltage scales with frequency 6V at 50 Hz, 30 V at 250 Hz.
Saturation voltage is determined by cross sectional area of core and number of turns - the product of core area and turns will be the same for any winding rated at 6V/50Hz. The smaller the core, the greater the number of turns required. This also means that, for the same voltage rating, smaller cores have higher primary inductance because inductance scales as N^2. So small cores good.
Yes, if amp is current protected, 1 ohm should be just fine.
R
No, I mean 1 H inductance. This is the impedance that appears in parallel with the reflected load impedance. This tells you how much current the amplifier must provide to generate magnetic field in the core, in addition to the reflected load current.
Yes, distortion falls with frequency 0.01% at 500 Hz, etc...
No you are not limited to 30 Vac. The audio signal comprises many different signals at different frequencies, its the lowest frequencies that determine whether it saturates or not- Which is why it is important that the crossover removes lower frequencies. Remember saturation voltage scales with frequency 6V at 50 Hz, 30 V at 250 Hz.
Saturation voltage is determined by cross sectional area of core and number of turns - the product of core area and turns will be the same for any winding rated at 6V/50Hz. The smaller the core, the greater the number of turns required. This also means that, for the same voltage rating, smaller cores have higher primary inductance because inductance scales as N^2. So small cores good.
Yes, if amp is current protected, 1 ohm should be just fine.
R
So.... you wire up the cells to the transformers and then:
1. you want to establish the impedance seen by your amp. You do that the usual way. I run REW curves with a few stock resistors which show me baselines; then I do a run with the transformers in place. You do not want your amp to see loads under 2 Ohms which is quite possible. So you might well end up with one Ohm in series with the transformer.
2. Then you need a mic to do frequency runs in order to see what EQ elements (resistance or impedance pieces) you want in replacement for the one Ohm or together with that one Ohm resistor. I found this step to be nerve-wracking as you are trying to maintain a sane load for the amp while doing what you can to flatten the acoustic output.
Ben
1. you want to establish the impedance seen by your amp. You do that the usual way. I run REW curves with a few stock resistors which show me baselines; then I do a run with the transformers in place. You do not want your amp to see loads under 2 Ohms which is quite possible. So you might well end up with one Ohm in series with the transformer.
2. Then you need a mic to do frequency runs in order to see what EQ elements (resistance or impedance pieces) you want in replacement for the one Ohm or together with that one Ohm resistor. I found this step to be nerve-wracking as you are trying to maintain a sane load for the amp while doing what you can to flatten the acoustic output.
Ben
Yes, that is the general approach taken with most hybrid ESLs.
You would only need to “tune” the Fs is you were still thinking about trying to use the resonance to equalize the low end roll-off due to having a finite-line source.
The spreadsheet doesn’t know anything about wires or spacing of wires or even if you are building with wires. It only knows that the segments will be ~1.2” wide with half of that width placed symmetrically about the central 0.6” wide segment (for Configuration 2)
The esl_seg_ui parameter inputs are calculated to model the width and number of your segments, and the ladder resistors since that is what matters for the acoustic response. 10 wires per segment was chosen arbitrarily and then the wire spacing input parameter calculated to get the correct segment width. So, for Configuration 2, you will always see 5,5,5,… as input for wire count. Configuration 1 will always have 10,5,5,…
Hopefully that makes sense now.
In my experience, you will only get flames if you use a highly conductive coating on the diaphragm. With round wire stators and a high resistance diaphragm coating you will likely only hear the dynamic peaks of the music get fuzzy sounding and start smelling ozone. If stators happen to have any sharp edges it is possible to get a stator-to-stator arc…you will see a little blue spark and notice later it has punched a small hole right thru the diaphragm. This is fairly common with perforated sheet metal stators. Surprisingly, panels can play on for years with many small arcing holes in the diaphragm without any noticeable change in sound quality.
Yes, I was wondering about that I was thinking the spreadsheet has no idea how many or what size wires I'm using 🙂.
My only concern, in my particular case, with this is that I had to cut the butt ends of the wires in the middle of the panel so they were lined up mostly in a row (a mistake I won't make again). So a couple of the ends are a bit rough (aka "sharp").
I will of course trying to deburr them (as such) as much as practical but I also plan to run a bead of the E6000 glue across them and put a piece of spacing tape over them. That really should be more than enough to prevent arcing to the diaphram.... I would think 🙂
Alright, after much starting/stopping and glue issues, I've got my 1st stator panel all segmented up and soldered together (finally).
Now that it done... I am questing my approach 🙂. Really thinking I should build a stretching jig and do it that way.....as I would probably be happier with it in the long run and will possibly dodge issues with rattling and arcing with the current approach.
I have a pretty good shop in my house and have access to relatives who are machinists (i.e. really cool tools) am I crazy to pivot now and do a stretched wire approach?
Now that it done... I am questing my approach 🙂. Really thinking I should build a stretching jig and do it that way.....as I would probably be happier with it in the long run and will possibly dodge issues with rattling and arcing with the current approach.
I have a pretty good shop in my house and have access to relatives who are machinists (i.e. really cool tools) am I crazy to pivot now and do a stretched wire approach?
I doubt you would have any rattling or arcing issues with the welding rods and they would sound about the same but stretched wire panels are definitely prettier--- and easier too, once you have a stretching jig.
Thanks...
The main factor is, the prospect of putting together another 6 tig rod "grids" is demoralizing me 🙂 (3x for the 36" then basically have to do another 3x ~12" for the top 1/3).... I am thinking my sanity and happiness would be much better off building a stretching jig 🙂.
Plus, though I have no science or data to back thus up, removing all of that plastic grillwork (louvers) just seems like a good idea (i.e. removing stuff from between me and the diaphram). Just have to wonder what the plastic grills is doing to the sound (diffraction etc...).
Not 100% committed to it yet but I may give it a shot. believe it or not I actually have some large pieces of 3" and 1 1/4" (.25" thick) aluminum angle laying around in my shop that will probably work or I could go the steel route.
I was thinking of doing a jig like this with the threaded rods. It is appealing to me to be able to make 2 stators at the same time 🙂.
http://www.diyaudio.com/forums/plan...segmented-wire-stator-esl-23.html#post4285619
The wood frame would be replaced by metal, I am actually kind surprised that jig didn't explode under the force of the wires 🙂.
I would love to get some more details about the jig(s)..... like where does the wire "anchor" when you start it; how the tension screws attach to the double stacked threaded rods to "pull" it; how to wind it such that you get all the wires wound on there "tight" such that you don't have some loose and some tight, to prevent the tight ones get more stretch than the loose ones etc....
My initial thought is use 20awg, that is somewhere around .05 to .06 ish O.D. Spaced at 11 TPI (5/8 rod). If I math'ed it correctly, it should keep me in the 40% to 50% open area range.
BTW, Charlie, where did you get your wire from?
On top of that, I just know in a year or so I will be building full size ESL panels. I have recognized long ago that the quest to build better and better audio gear is a progressively terminal disease 🙂.
The main factor is, the prospect of putting together another 6 tig rod "grids" is demoralizing me 🙂 (3x for the 36" then basically have to do another 3x ~12" for the top 1/3).... I am thinking my sanity and happiness would be much better off building a stretching jig 🙂.
Plus, though I have no science or data to back thus up, removing all of that plastic grillwork (louvers) just seems like a good idea (i.e. removing stuff from between me and the diaphram). Just have to wonder what the plastic grills is doing to the sound (diffraction etc...).
Not 100% committed to it yet but I may give it a shot. believe it or not I actually have some large pieces of 3" and 1 1/4" (.25" thick) aluminum angle laying around in my shop that will probably work or I could go the steel route.
I was thinking of doing a jig like this with the threaded rods. It is appealing to me to be able to make 2 stators at the same time 🙂.
http://www.diyaudio.com/forums/plan...segmented-wire-stator-esl-23.html#post4285619
The wood frame would be replaced by metal, I am actually kind surprised that jig didn't explode under the force of the wires 🙂.
I would love to get some more details about the jig(s)..... like where does the wire "anchor" when you start it; how the tension screws attach to the double stacked threaded rods to "pull" it; how to wind it such that you get all the wires wound on there "tight" such that you don't have some loose and some tight, to prevent the tight ones get more stretch than the loose ones etc....
My initial thought is use 20awg, that is somewhere around .05 to .06 ish O.D. Spaced at 11 TPI (5/8 rod). If I math'ed it correctly, it should keep me in the 40% to 50% open area range.
BTW, Charlie, where did you get your wire from?
On top of that, I just know in a year or so I will be building full size ESL panels. I have recognized long ago that the quest to build better and better audio gear is a progressively terminal disease 🙂.
Last edited:
I used single strand 20awg / .010XLPVC wire purchased from Interstate Wire.
This wire is .052" O.D. which spaces to 43% open if you use 12TPI threaded rods to space the wires.
My jig uses 1/16" diameter x .25 long steel pins set into 3/16 aluminum plates. I drilled the plates with short 1/16 bits. All items purchased from purchased from McMaster-Carr. The pins are set into the plates angled 4 degrees past vertical so that they hook the wire loops.
My jig is made from 4x4 pine and it's strong as hell. The jack screws are (2) 3/4 threaded rods with coupling nuts and large washers. The threaded rods are welded to a 3/16 steel plate. The aluminum plates are attached to the steel plate with six 1/4-24 countersunk screws.
My jig will only do one stator at a time though, assembling the wood lattices over the wires in the jig, but it works very well. I put down wax paper on the jig platform before stringing the wires-- the glue lines are perfectly flush to the wire surfaces.
So been reading through this a bunch of times the past few days. And I was wondering, for the stator frame for the wires, is there a recommend "height" for the spacers?
http://www.diyaudio.com/forums/plan...segmented-wire-stator-esl-23.html#post4285619
i.e. how "tall" should the ribs of the frame be? (rule of thumb, if any).
Think I got a plan together to build the "threaded rod" type stretcher jig together. Going to try to put it together this weekend. The notable difference is, mine will be made of steel angle or square tubing 🙂
http://www.diyaudio.com/forums/plan...segmented-wire-stator-esl-23.html#post4285619
i.e. how "tall" should the ribs of the frame be? (rule of thumb, if any).
Think I got a plan together to build the "threaded rod" type stretcher jig together. Going to try to put it together this weekend. The notable difference is, mine will be made of steel angle or square tubing 🙂
Last edited:
My panel configuration was a special case because it had to fit into my existing speaker frame (which was built for a 12x48 perf-metal panel with only 3/4" allowance at the edges)-- and it had to be thin enough to insert into the frame and clear the beam splitter behind it. Otherwise I would have made the horizontal ribs a bit taller.
The horizontal ribs in my oak support lattice are 3/16" thick x 11/16" tall x 12" wide and the open span between ribs is about 2 1/8". I think Bolserst or Golfnut posted some guidelines on the forum somewhere regarding wire size versus minimum span between ribs but you'd have to search for it.
If I had more clearance to the beam splitter I would have made the ribs at least 3/4" tall-- maybe 7/8", to restrain resonances. And in hind-sight I would have added one more rib to close the span between them a bit. Still, the panels sound really good as-is.
Last edited:
Cool, I was thinking something like 3/4" as they would be recessed fully into a 1.5" thick frame.
In the link I referenced, bolserst suggested this, which is what I am working under (though I will probably use less than 3").
"Concerning crossbar spacing:
It is important to consider crossbar spacing when selecting wire size.
The thicker the wire, the further apart you can space the crossbars. Here are some recommendations based on building experience:
22 AWG, Crossbar spacing = 2” max
20 AWG, Crossbar spacing = 3” max
18 AWG, Crossbar spacing = 4” max "
In the link I referenced, bolserst suggested this, which is what I am working under (though I will probably use less than 3").
"Concerning crossbar spacing:
It is important to consider crossbar spacing when selecting wire size.
The thicker the wire, the further apart you can space the crossbars. Here are some recommendations based on building experience:
22 AWG, Crossbar spacing = 2” max
20 AWG, Crossbar spacing = 3” max
18 AWG, Crossbar spacing = 4” max "
Is this a true danger using metallized mylar as a diaphragm ? I am building a wire segmented stator using pvc insulated copper wire and have purchased some emergency blankets (12 micron aluminized mylar) for diaphragm usage. I have used it before without problems but will not use it again if there is only the slightest threat of setting my house on fire 😱
I have read varying comments on other disadvantages of metallized mylar including charge migration and distortion. However , there are some commercial esl's that use them and I have a read a book ( Dutch, E Fikier) about DIY esl's which descibes various designs solely using metallized mylar.
Not having to apply custom coating is the main advantage for me.
The licron spray on mylar isn't too bad... a bit expensive for the spray I suppose.